Abstract:Traffic sign recognition is a technology by which a vehicle is able to recognize the traffic signs on the road. In this paper, we propose a novel traffic sign recognition that can operate robustly and accurately for real scenes of Korean roads. The proposed method first detects a potential traffic sign and then recognizes the content of the potential traffic sign. With this approach, the proposed method can robustly recognize small traffic signs from long distances and reduce false alarm significantly. We employ Histogram of Oriented Gradient (HOG) and Support Vector Machine (SVM) in a two-step model. Compared with the original HOG and SVM method using three Hyundai data sequences with ground truth, our proposed method outperforms significantly and operates robustly in different conditions. The source code and datasets are available online at https://github.com/comvisdinh/realtimetrafficsignrecognition.
Keywords: Traffic sign recognition, histogram of oriented gradients, support vector machine.
|
DOI:
10.17148/IJARCCE.2022.111126
